Abstract
Assessments of the impacts of ozone (O 3 ) on regional and global food production are currently based on results from experiments using open‐top chambers ( OTC s). However, there are concerns that these impact estimates might be biased due to the environmental artifacts imposed by this enclosure system. In this study, we collated O 3 exposure and yield data for three major crop species—wheat, rice, and soybean—for which O 3 experiments have been conducted with OTC s as well as the ecologically more realistic free‐air O 3 elevation (O 3 ‐ FACE ) exposure system; both within the same cultivation region and country. For all three crops, we found that the sensitivity of crop yield to the O 3 metric AOT 40 (accumulated hourly O 3 exposure above a cut‐off threshold concentration of 40 ppb) significantly differed between OTC and O 3 ‐ FACE experiments. In wheat and rice, O 3 sensitivity was higher in O 3 ‐ FACE than OTC experiments, while the opposite was the case for soybean. In all three crops, these differences could be linked to factors influencing stomatal conductance (manipulation of water inputs, passive chamber warming, and cultivar differences in gas exchange). Our study thus highlights the importance of accounting for factors that control stomatal O 3 flux when applying experimental data to assess O 3 impacts on crops at large spatial scales.